Microsoft completes GitHub acquisition
blog.github.com
blog.github.com
Microsoft could pull a Skype of course, but on the upside this might turn out to be a good thing, as in this climate many businesses, operating at scale and giving away so many freebies, are struggling and GitHub could have been the next SourceForge.
Who knows, maybe they'll even open source it. Fingers crossed.
A better example would be Nokia.
In my industry, Microsoft stacks are extremely common. The talent pool for C# in this area is tremendous, which has absolutely gained the company much synergy in the form of OS and MSSQL licensing. So the development of .Net in my opinion has made the company much more sticky.
I can't help but believe they're making a mistake (even though I want it!) in pushing .Net to open source.
If Microsoft can't retain developer mindshare, they are sunk.
Azure is a huge step forward in bringing many of these traditionally in house technologies like SQL Server, SSIS, SSAS, Reporting Services, and .NET Web Apps affordably into the cloud as well as newer technologies that fit comfortably into companies' existing infrastructure (Data Factory, Azure Functions, HDInsight etc.)
They are grabbing a huge amount of developer mindshare at the enterprise level, and by opening up their tools are allowing the younger crowd a free alternative to going with OSS.
I'm definitely not a MSFT zealot. I have been a "linux guy" since I was 14 yrs old, but after working almost exclusively with the MSFT stack for the last 2.5 yrs I am very impressed with their offerings and the level of integration and cohesiveness around those offerings.
I would have to disagree. .NET has become the backbone of the MSFT developer experience and has gone far beyond being just an alternative to java. It's heavily used in Azure, SQL Server, Web Applications, Desktop, Mobile etc.
eh? the whole reason they make tools is to lure developers.
Yup. Very glad that can never happen now.
Microsoft won't ruin GitHub. Windows's ecosystem need developers more than ever. Developers are on GitHub. They simply can't screw it up.
I think Microsoft has done this a lot in the past. Mobile is the standout example, but even the Skype debacle had some good intentions. Microsoft had a super fragmented ecosystem of "Windows Live", Office, MSN, Skype.. I think it was noble to try and unify everything under one login.
But they tried to push said login down the throats of unsuspecting Windows users (that more often that not didn't use that MS ecosystem but instead used Google's), and (IMO) ruined Skype's usability in the process.
I don't know what the _right_ way of doing it was, but this certainly wasn't it.
WSL is not perfect, but it's good enough for a lot of use cases if you're stuck on Windows.
When will MS stop taking money away from the developers by making Visual Studio priced with hefty charges?
I still can't believe they're just making a barrier to entry by charging people who are going to help their ecosystem better and bigger.
I still like Skype more than Discord, but chat apps are wherever your friends are.
What kind of world do we live in when a service which is so ubiquitous and depended upon in our community is at risk of struggling?
(Side note: I haven't been following this acquisition in detail; was GH struggling? Or was the acquisition just their exit?)
Github would be unknown today if it wasn’t free for personal use.
> Ultimately, my job is to make GitHub better for you.
Whenever I see a product claim it's "for you", I cringe. You just can't put something on the web or on TV and say "for you" and have it mean anything useful.
What I always loved about GitHub was that it was the underdog, so they naturally built tools so that a couple of developers in a garage (well, cafe) could have access to the same type of tools that a Fortune-100 company has (and has to pay companies like IBM big bucks to install and maintain). It's using technology to help those with fewer resources, which is to me the entire purpose of technology.
Almost everything I've seen from GitHub in the past year or two has been to help enterprise developers at big established companies. Creating giant workflows and integrating with legacy systems and such. The little things that individuals use are slipping through the cracks. When I reported that the milestone date-picker went away for Safari users, for example, they just told me "Sorry".
"Developers" has become a dirty word. It's a weasel word that companies use to try to convince me they're talking about me, while actually talking about someone and something completely different. If you can't be more specific than that, you're almost certainly wasting my time.
I miss the GitHub whose homepage had logos of little startups I'd barely heard of, and direct links to new and interesting repos. Now they've got IBM and SAP and Walmart logos. It's clear that the "developers" they're targeting no longer includes me.
People used to say, "If you're not paying, you're the product", so I've been careful not to depend on any free-with-a-catch services. I guess now I have to be extra careful not to depend on any VC-funded services, either, because if I'm not as rich an income stream as the corporate world, I'm still not the real customer.
With that said, tech capitalism is hard! Even if you pay, as you mention, you might still end up being left in a lurch.
I don't know what the solution is unless you build everything yourself, which is untenable, or ensure that providers you use are easy from when necessary.
When private fire departments fail to work well for citizens, we no longer just say "Well, fire department capitalism is hard!" It's the wrong model.
Companies eventually go after companies willing to spend money, otherwise they don’t survive because consumers and startups are cheap and don’t want to spend money on such productivity tools and services.
That or they could’ve shoved ads down your throat, or bundle spyware in the archives people download.
As they say, damned if you do, damned if you don’t. Next time when you like a service that brings you value, pay for it (using the royal you here).
As it is, they're moving in a direction that's getting worse for me over time, so "get all my (paid) stuff off GitHub" is on my to-do list.
If you're suggesting that every company that starts with consumer-oriented products must inevitably transition into enterprise-oriented products, I don't believe that to be true. I can think of many counterexamples.
Instead they insisted on endless growth, hosting All The Projects without bound or limit and I’d say that’s what puts them in hot water, needing crazy amounts of money that individuals can’t provide.
$8.4 million in gross revenue a year. Out of this has to come all operational costs, all salaries, all benefits for staff, all other overhead, and so on. At west coast tech rates, that's low-30s number of people. Assume lower, since AWS and similar isn't exactly free. Do you think GitHub could have done every single thing they did at the time you have in mind with a double handful of engineers and sysadmins and managers? I must admit, I'm a little skeptical. Staffing a solid ops organization capable of 24/7 coverage takes at least a dozen people.
I haven't even touched on taxes.
Business plans can easily be a thousand a month or more. GitHub has an offering of $21 per user per month, also for business. In either case this almost certainly offers significantly greater overhead that the individual plans do. You can get the same revenue numbers with significantly lower load with a third the number of users! Not to mention that you have fewer customers for the same amount of revenue, which makes support higher-touch but much easier in important ways.
With this in mind, it's easy to see why a group like GitHub might want to consider investing in supporting businesses. There's so much more margin there. Especially when you consider that their predecessors did not, and GitHub out-competed them in part due to the extra revenue that comes from strong enterprise support.
Services like GitHub don't get to sit still. Or else we'd all still be using SourceForge.
Personally, from GitHub, I didn’t want 9 9s or infinite bandwidth or feature creep or even to use it at my day job (Jira/Jenkins/prayer).
SourceForge is a scarecrow argument btw, it’s the MySpace of its sort.
Sourceforge is exactly like MySpace! Sourceforge once ruled the roost. It was, for many people, the only game in town. It got complacent, and thought it could stay top dog while only really caring about the not-particularly-profitable users or if there was a good way to make more money from the services it offered.
Competitors innovated, found other business models, and did much better. Now we have GitHub! Which has learned from the missteps of Sourceforge. One of which was failing to invest in features that high-margin customers value and will happily pay extra for.
My original comment was an argument that the reorientation towards business clients was not an inevitability as the parent comment said it was. this mindset says products for rational, aware individuals (here, developers) are not worth building anymore since the margins aren’t high enough. If these individuals aren’t worthwhile clients anymore than they are only data to be traded.
There is a very reasonable point that this should be enough for any company not addicted to endless, pointless growth. This is, after all, the ideology of a cancer! Customers should be respected, rather than treated as data-generators fit only to be traded upon. You have made this point wisely and well.
I think it's worth considering why a business might consider limiting its investments in competing for a pool of low-margin customers of limited size. Historical examples suggest that confining your investments to just this customer pool will often end with your customers being attracted away to other providers who have the resources to produce a superior offering. As a result, what should be a healthy and sustainable mindset of building for rational, aware individual consumers can easily lead to being out-innovated and out-competed.
An intelligent reader will note at this point that there is a lot of uncertainty in the previous paragraph. This person is right! There is a great deal of uncertainty at hand, as with all things in markets. With that said, the above scenari is more likely than not, and the expected gains from chasing high-margin customers are generally larger than the expected gains of opting to focus exclusively on low-margin ones in the face of competition.
To such services as GitHub, the consumer offering is not a way to collect data to be monetized. Indeed, the experience for single consumers is critical - it shows the key features and accustoms users to the product. A high-quality offering for rational, aware individual consumers is of paramount importance because it is where the real customer funnel starts.
The first job I had, I needed to get a visual studio licence of some specific level. It was a $2000 dollar licence or something. The bureaucracy of the organization I was in was large and slow.
My manager told me to request 4 licences instead of 1, so he wouldn't have to do the paperwork for the next set of interns. This wasn't an under the table thing, he told me this in front of /his/ manager, who obviously agreed that this was a sensible decision.
As another example, my current employers attitude towards s3 space usage is "eh, keep shoveling it in there, it's cheaper to let it grow indefinitely than pay you to make something that clears it out automatically. We'll let you know if it ever gets expensive.". I can't say I fault the logic.
Customers are certainly willing to pay money, but business services are where the gravy is at.
And yes I’m familiar with expensive licenses, we have one which is $20k per physical dongle and it’s terrible.
[1] https://github.com/isaacs/github/issues?q=is%3Aissue+is%3Aop...
Plus I'm guessing many/most GH employees aren't looking at this, this kind of listings is generally a bummer/PITA, and fixing them is a thankless job: once you've done so, all you get is "finally, took you long enough".
This is the paradox of GitHub. The product itself is closed-source, so the outsiders cannot do the work to implement that feature themselves, no matter how much they want it.
For the same reason I use a Gnu/Linux operating system. Truthfully there are things that I imagine MS or apple might do better, but that's beside the point for me. It's a political stance.
FOSS development is way too much concentrated on GitHub already, it's unhealthy. But there is no problem in most people not thinking like you either.
Overall, I think the acquisition was good, if no other reason, just because there are enough people that dislike MS for pushing some diversity on the FOSS hosting game.
Microsoft feels like an untrustworthy friend to me -- I'll go to a party at Microsoft's house, but I wouldn't let Microsoft hold my wallet.
I really like how Microsoft is now, all open, free, unicorns and rainbows. Hell, they're producing some terrific open source themselves and support community.
But lets not forget - corporations are not people. Next CEO may use Github to put obstacles on OSS movement's way. Just because it is good now, doesn't mean it will be good tomorrow, or in 10 years.
P.S. Personally, I really happy where Microsoft is heading right now, and I think they're doing terrific job with their open source projects. You go Microsoft. I hope OSS way would be successful enough so MS would embrace it even more... But you never know.
EDIT: [1] I'm aware that Microsoft recently joined Open Invention Network patent pool. But there're criticism that contributed patents are not important ones. I tried to find ExFAT patents, but couldn't in OIN list. Also, it is really hard to find full OIN contributed patent list. Here's US patent numbers MS going around asking money for, couldn't find any confirmation of all of them being contributed 5579517 5758352 5745902 6286013
Not the op, but yes, both "Corporations are not humans" and "Corporation's acts are driven by profits" also applies to RedHat and other publicly traded companies.
You can either fear the future and hence never enjoy what's good today. Or you can forget the future and actually enjoy the present.
GitHub didn't exist 15 years ago. Something better will probably be there in 15 years from now.
If GitHub changes direction and becomes a bad citizen, we'll go elsewhere.
Unless they buy and shut down any alternative that gathers more than a thousand users.
A better choice would be to enjoy what's good today while making contingencies for when things go south in the future.
With .NET Core [0], and ChakraCore[1] (Edge's JS engine) both being open sourced. And them having thrown the entirety of MS-DOS [2] up under the MIT license, and making a bunch of new tools and stuff under open licenses (VSCode, TypeScript, msbuild, etc...). I honestly wouldn't be surprised if they started making a push to open up many parts of Windows.
Already some devs on the Edge team have said publicly that they are working on open sourcing the whole rendering engine from Edge, and seeing that they are moving toward having Windows be more of a "platform" and less of "software you buy", it makes sense that it could eventually become an open source (or at least "open core") product.
I don't think it will happen this year, but I do think that if MS keeps going in this direction, it's more likely than not to happen eventually.
[0] https://github.com/dotnet/core [1] https://github.com/Microsoft/ChakraCore [2] https://github.com/Microsoft/MS-DOS
I think MS in the past few years has made huge strides...
I'd far rather they look to open source individual components one at a time. Most of the modern UWP apps of theirs they could easily open source, for instance. Individual components of Windows may be reasonably open sourcing them as they go through and upgrade and enhance them.
Did MS confirm that them joining OIN includes non aggression with exFAT and ActiveSync patents?
One could even say that GitHub was "embracing & extending" Git (pull requests, issue trackers, wikis) well before the acquisition.
One way to look at is that code is no longer that interesting, data is more important.
Does any of those share their data under an open license?
Having it in RackSpace and AWS is bad optics, but they also must know Azure is not stable enough for something as large and visible as GitHub.
I expect this to be the first integration they try. Stockpiling the popcorn.
In what universe is this true? Azure is the backbone of large things in corporate America, especially in banking and healthcare that you never even think about. Things a lot more important than someone's repo being inconvenienced.
Any migration will be painful, but to say Azure as a service isn't up to the task is just nonsense.
For the sectors you cited that might not matter since those will not be using cloud resources but rather the common virtual machine workload.
[0] https://about.gitlab.com/2018/04/05/gke-gitlab-integration/
But another issue with this comment is this:
> in RackSpace and AWS
As far as I can tell, they moved off RackSpace a long time ago. And they only use AWS infrastructure for peering, which makes sense with the amount of AWS machines that pull code from Github. The majority of the heavy lifting is done in Github's own data centers: https://githubengineering.com/evolution-of-our-data-centers/
Who asked you to make GitHub better though? It's pretty darn good already. I fear that the new CEO will try to make things better when they are already good. I'm really scared that GitHub will end up like Azure DevOps, perhaps get merged or shut down entirely when they feel they have feature parity. Currently there is no great GitHub alternative IMHO and for selfish reasons I wish Microsoft could just leave it. The most I am scared is to wake up one day and see the most ugly "Metro" UX applied to GitHub.
Alongside this, is there also a way to change the order of Wiki Pages?
I think you can override the default index of pages with a custom one.
This was my draw to it as well.
GitHub will retain its product philosophy. We love GitHub because of the deep care and thoughtfulness that goes into every facet of the developer’s experience. I understand and respect this, and know that we will continue to build tasteful, snappy, polished tools that developers love.
Of course they will make big changes two years later down the line. You don't buy something if you have no plans for it.
Well, _technically_ I worked for one where it functioned in a hands-off way. The big difference was, the parent company was a holding company specialising in a certain industry, and us functioning differently from their other properties was the whole point.
In this case, I expect things to go the What's App way: it's going to be long and slow, but MS will eat it eventually.
GitLab is decent, with an open source core that can be installed on premise — I introduced it for a former employer and they are still very happy with it.
All isn't perfect and I'm still on GitHub due to the network effects, but you do have alternatives.
If you have more feedback on the things that could be better, we would love to hear more!
Since the news broke I've been hosting a Gitea instance and while that works fine, I cannot go a day without using Github for things like pulling other people's code, creating issues, etc.
Seriously it only takes a bunch of high profile projects and people to drive a migration.
In other words Microsoft has to play nice.
SourceForge is from before my time. When Github was still small, I was still being brainwashed by school's free Windows and discount Office licenses and school educating us in using Microsoft Office. From my current point of view, though, SourceForge was never as big as Github is now. Almost every development company pulls code from Github, either in an automated fashion to pull in updates, or casually as they need random things that are published there. I don't think that was ever the case with SF.
If you're good with Powershell it's also extremely easy to write your own build and/or release steps which is a huge plus.
Just a few days ago the service connection to Azure suddenly stopped working with "401 CouldNotFetchAccessTokenForAzureStatusCode" which appeared when deploying to Azure AppService. We had to recreate the service connection to get it working again.
Recently we also wanted to move our Android builds to Azure DevOps but in the end the documentation was so confusing that we just used Bitrise instead.
How hard is to bring TFS in house?
It is? Because it takes my home feed ~5 seconds to load. Totally unacceptable.
Microsoft, his employer and the owner of GitHub.